Objectives Although physical activity can potentially reduce symptoms of arthritis, 50% of people with arthritis are insufficiently active. The aim was to identify psychosocial factors associated with increased physical activity in mid-age adults with arthritis who did not meet recommended physical activity levels.

Design Longitudinal cohort study.

Methods Data were from 692 insufficiently active men and women (mean age 55 ± 6.6 years) with arthritis, who answered mailed surveys in 2007 and 2009 in the HABITAT study. Increased physical activity was defined as a change of ≥200 MET min/week in walking, moderate and vigorous activities from 2007 to 2009. Scale scores were used to measure psychosocial factors including intention, experiences, attitudes, efficacy, barriers, motivation, social support, and health professional advice. Associations between (1) 2007 psychosocial factors and (2) 2007–2009 improvement (≥+1 standard deviation) in psychosocial factors and increased physical activity were examined with logistic regression models. Results were adjusted for education, body mass index, and self-rated health.

Results Between 2007 and 2009, 296 participants (42.8%) increased their physical activity. Engagement, mastery and physical activity intention in 2007 were associated with this increase in physical activity (engagement OR = 1.11, 99% confidence interval (CI) = 1.05–1.17; mastery OR = 1.12, 99%CI = 1.02–1.22; physical activity intention OR = 1.29, 99%CI = 1.06–1.56). Improved scores for encouragement (OR = 2.07, CI = 1.07–4.01) and self-efficacy (OR = 2.27, CI = 1.30–3.97) were also significantly associated with increased physical activity.

Conclusions Positive physical activity experiences and intentions were predictors of increased physical activity among people with arthritis. Improved physical activity confidence and social support were associated with increased physical activity. It is important to consider these psychosocial factors when planning physical activity interventions for people with arthritis.